Modeling Passing Maneuver Based on Vehicle Characteristics for In-Vehicle Collision Warning Systems on Two-Lane Highways

Abstract

Modern vehicles are equipped with various sensors of high accuracy and sensitivity, based on which it is possible to implement passing collision warning systems (PCWS) for two-lane highways. In previous systems, the time required to complete the passing maneuver safely was formulated based on pre-established regression models. In this paper, this time is formulated based on actual vehicle characteristics. The new vehicle dynamics model for the PCWS prototype includes steering control and drivetrain models, and allows more accurate prediction of the required passing time. The geometry of the passing maneuver (for the case of an impeding truck), the main phases of the passing process, and the distances related to the conditions for predicting passing time are described. The interactions between the PCWS and driver actions are formulated. The steering control model is based on a two-dimensional perspective representation of the three-dimensional reality perceived by the driver. The drivetrain model, including inertial and mechanical losses in the drivetrain, considers automatic gear shift and the presence of a torque converter to simulate vehicle performance accurately. The proposed PCWS was tested using MATLAB Simulink.